[ 
https://issues.apache.org/jira/browse/HDFS-9607?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15146636#comment-15146636
 ] 

Dinesh S. Atreya commented on HDFS-9607:
----------------------------------------

Konstantin,

Agree with your statement:
{quote}
Dinesh, this is an application-level logic.
{quote}
It will be good to indicate at the API level that HDFS will carry out 
{{canWrite}} checks with the information it has or it has been provided with. 

For example in the initial implementation we can use {{canWrite}} checks to 
disallow updating information more than a HDFS block length, certainly beyond 
the end of a file (even beyond the end of a block). Higher levels, for example 
applications then can override this method to carry out additional checks. (_I 
will log JIRAS for other projects so that they can provide inputs/perspectives 
including additional requirements_).

In future {{blockWrite}} API may be introduced for updating info spanning 
multiple blocks as needed. (*_Separately I need to create another JIRA for 
block level leases in addition to file level leases that is prevalent now._*)

> Advance Hadoop Architecture (AHA) - HDFS Update (write-in-place)
> ----------------------------------------------------------------
>
>                 Key: HDFS-9607
>                 URL: https://issues.apache.org/jira/browse/HDFS-9607
>             Project: Hadoop HDFS
>          Issue Type: New Feature
>            Reporter: Dinesh S. Atreya
>
> Link to Umbrella JIRA
> https://issues.apache.org/jira/browse/HADOOP-12620 
> Provide capability to carry out in-place writes/updates. Only writes in-place 
> are supported where the existing length does not change.
> For example, "Hello World" can be replaced by "Hello HDFS!"
> See 
> https://issues.apache.org/jira/browse/HADOOP-12620?focusedCommentId=15046300&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-15046300
>  for more details.
> Currently following are supported.
> # Sequential writes
> # Append (HADOOP-1700, HDFS-265)
> # Snapshots (HDFS-2802)
> # Truncate (HDFS-3107)
> This JIRA is for random updates (write-in-place).



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to